I have a question.
how I can change the column in the gc/ms without stopped turbopump ? It is possible?
regards Azrael ?
I believe what @mhardacker is referencing with the flow calculation is that as soon as the restriction capillary is cut, the EPC is no longer controlling the amount of flow to the MSD, i.e. atmospheric pressure…
There is a product for this.
Use 1 meter of deactivated fused silica tubing going into the source (the same diameter of your column), then use a union (either the press-fit glass type of, better yet, the SilTite brand) to connect to your column in the oven. When you want to change columns, cut the fused silica just after the union and plug it with an old septum while you do the change. Then affix the new column to the union and finally remove the old septum, cut a few mm of the fused silica to remove the rubber plug, clean and affix to the union. Done. I have been doing this for 30 years and it has never failed me.
Or you can go the expensive way and purchase the Agilent Quick Swap attachment. I have no experience with this, though.
The above statements are correct & I would like to point out here from my experience as a service engineer:
1a) Azrael is probably responsible for several users
1b) many "modern" users do not come with our practical experience.
2) Pressfits are good, but in my opinion not suitable for less experienced users or GC novices =>
then metal ferrules like AT CFT (Capillary Flow Technology) or Siltite are the better/more reliable choice.
3) Short MSD restriction capillary: can I use any dimension? Please, NO
Why are the dimensions of the MS restriction capillary so important?
When the main column is removed, only the restriction capillary (inside the MSD) limits the air flow into the MSD/pump.
The dimensions of this capillary (iD & L) should be calculated, because more than 6(-10)ml into the MSD/pump should be avoided !
Examples for bad & good AIR flow values by restriction capillaries (no gas supply by EPC or main column):
Dimensions of restriction capillary [ID µm * L]
Air Flow into MSD,
(without main column)
Idea: where & who to use?
(not mandatory to do)
BAD, too high flow
250 * 1.0 m
250 * 2.5 m
if main column has ID 250µm.
extend the main column by 2.5m
200 * 1.0 m
if main column has ID 200µm.
extend the main column by 1 m
Calculation details: see screenshots below
Because of all these mentioned uncertainties I always recommend the Agilent QuickSwap.
Admittedly, QuickSwap might be not cheap (because of EPC) but in the environment of many users it’s the SAFEST hardware choice and it also protects expensive equipment very well.
Note: when changing columns the ion source/quad temperatures should be reduced to 100°C to avoid unnecessary oxidation.
Good points. I agree that the quick-swap is the best solution if the expense can be justified. As to press-fit vs metal ferrules, it's a hard decision. They both require care on the part of the person doing the column swap, and the ferrules cost twice as much as the press-fit. One has to find what works best for the situation.
I do have one comment in regards to item 3:
I have never considered that final length of capillary to be a restrictor, but an extension to the column (making sure that it's the same ID as the column). Then I just add the length of the capillary to the length of the column in the column definition and the EPC takes care of keeping the flow rate at the desired value. Not that it will make a major difference in any production environment. As long as you get the chromatographic resolution and run length that you need/want, the actual flow rate should be irrelevant as long as it's within the instrument's capabilities. The 14.5 mL/min are the maximum that the 1 m length of 0.25 mm capillary can transfer, but that volume isn't available because the EPC is controlling the flow from the column, so it's only that much volume that gets into the MS.
I believe what @mhardacker is referencing with the flow calculation is that as soon as the restriction capillary is cut, the EPC is no longer controlling the amount of flow to the MSD, i.e. atmospheric pressure is acting as the EPC supplying about 15psi to the front of the 1m x 0.25mm restriction capillary and at 30C, that will allow about 15mL/min of atmosphere into the MSD. As long as that capillary is open to the atmosphere for only a short time, then one would assume it is a minor issue to apply such a large gas load to the MSD's vacuum system. In my opinion the concerns would be: how much atmosphere enters the system (oxidation) and for how long (which may cause the user to wait some time before use to pump out the atmosphere), repeated loading and unloading of the turbo pump due to large changes in flow to the MSD (relative to the 'normal' column flow during analysis) caused by plugging and unplugging the restriction capillary with a septum and connection to different sources of gas (EPC with helium versus atmosphere).
The turbo pump is spinning very fast under normal operating conditions and it is designed to pump at a certain rate which in the case of the MSD is assumed to be a reasonably low and stable amount of gas (load), see excerpt from the 5975C MSD data sheet below. What is not clear are the affects of rapidly changing the amount of gas to which the turbo pump is exposed >>someone with more knowledge about materials and the dynamics of changing forces would need to chime in here<<. For novice users one should be concerned with a catastrophic event whereby the restriction capillary cannot be reconnected quickly to whatever column union device is being used. Allowing the restriction capillary to be open to the atmosphere for an extended period is likely to result in the MSD system venting itself as a failsafe due to the high flow. A possible preventative measure would be to increase the length or reduce the diameter of the restriction capillary such that the maximum possible atmospheric flow rate is much lower. Another would be to reduce the temperatures of the MSD source and quadrupole to reduce the likelihood of oxidation prior to the column change.
To address all the concerns, Agilent recommends venting your MSD as well as cooling any heated zones connected to the column(s) with carrier gas still flowing. Once the MSD is vented and the affected zones are cool, mitigate the GC's failsafe of shutting down the inlet (or EPC) supplying gas to the column being changed (this is best done by turning the pressure setpoint 'off' for the affected EPC, either in the software or from the GC itself). The loss of pressure at the inlet or the EPC by removing the column will be seen by the GC as an unmet setpoint, so turning the setpoint 'off' [do not change the setpoint to be zero since the EPC modules are not designed to maintain a setpoint of zero] will tell the GC to not worry about controlling the pressure, no matter what it is. Once the column change is completed, turn the pressure back on so the GC will try to attain and maintain the previous setpoint to help decide if there is now a leak. Once the carrier flow has been reestablished for a few minutes to allow the column to have a few volumes of carrier gas flow through it, one can begin heating the zones back up. Refer to your SoP for specifics, but generically one needs to confirm that the leak rate (if any) is acceptable, then confirm function has been restored. Agilent recommends waiting 2 hours at operating temperatures to establish thermal equilibrium so repeatable results can be obtained.
A quick swap device or similar addresses many of these concerns, but as with anything, there are tradeoffs.
Ah, I didn't think of that. Yes, it makes sense...
That being said, after 30 years of doing it this way I haven't hit any problems that I could trace to having the short capillary open to the atmosphere for the time that it takes to make the changes. Depending on the ability of the person doing it, this can range from some 30 seconds to several minutes. It is likely that some oxidation might occur. But in a production environment (in my case, commercial environmental testing labs), one has to clean the source quite often due to the amount of "dirt material" that gets injected into the system, that perhaps that's why I haven't noticed ill effects. We clean sources once a month on the average (about 800-1000 sample extract injections, not counting standards or instrument QC). Sometimes more often.
But you are absolutely right that when dealing with inexperienced users, one should make it as safe and easy as possible, and you can't beat the quick-swap for that.